Journal of Pediatric Psychology Advance Access originally published online on January 9, 2006
Journal of Pediatric Psychology 2006 31(9):928-938; doi:10.1093/jpepsy/jsj098
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Effects of Behavioral Family Systems Therapy for Diabetes on Adolescents’ Family Relationships, Treatment Adherence, and Metabolic Control
1 Nemours Children’s Clinic, and 2 Washington University, St. Louis School of Medicine
All correspondence concerning this article should be addressed to Tim Wysocki, PhD, Center for Pediatric Psychology Research, Nemours Children’s Clinic, 807 Children’s Way, Jacksonville, Florida 32207. E-mail: twysocki{at}nemours.org.
Received May 10, 2005; revision received July 5, 2005 and August 1, 2005; accepted December 11, 2005
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionAcknowledgmentsReferences |
|---|
Background Behavioral family systems therapy (BFST) for adolescents with diabetes has improved family relationships and communication, but effects on adherence and metabolic control were weak. We evaluated a revised intervention, BFST for diabetes (BFST-D). Methods One hundred and four families were randomized to standard care (SC) or to 12 sessions of either an educational support group (ES) or a BFST-D over 6 months. Family relationships, adherence, glycosylated hemoglobin (HbA1c), and health care utilization were measured at baseline and after treatment. Results BFST-D significantly improved family conflict and adherence compared to SC and ES, especially among those with baseline HbA1c
9.0%. BFST-D and ES significantly improved HbA1c compared to SC among those with baseline HbA1c 9.0%. Conclusions The revised intervention (BFST-D) improved family conflict and treatment adherence significantly, while both ES and BFST-D reduced HbA1c significantly, particularly among adolescents with poor metabolic control. Clinical translation of BFST-D requires further study. Key words: adolescence; diabetes; family therapy.
Family conflict, parent–adolescent communication, and problem-solving skills have been associated with diabetes outcomes among adolescents in cross-sectional (Anderson, Miller, Auslander, & Santiago, 1981
; Bobrow, AvRuskin & Siller, 1985; Miller-Johnson et al., 1994; Wysocki, 1993) and prospective (Gustafsson, Cederblad, Ludvigsson & Lundin, 1987; Hauser et al., 1990; Koski, Ahlas & Kumento, 1976) studies. Thus, the development and validation of behavioral and psychological interventions that improve family communication, problem solving, and conflict resolution is warranted and could enhance diabetes management among adolescents and their families.
Behavioral family systems therapy (BFST) is a flexible, multicomponent, family-focused intervention that targets family communication and problem solving, extreme beliefs of parents and adolescents that impede communication, and systemic barriers to problem solving (Robin & Foster, 1989). The effectiveness of BFST has been verified in several studies (Barkley, Guevremont, Anastopoulos, & Fletcher, 1992; Foster, Prinz, & O’Leary, 1983; Robin, Seigel, Kopeke, Moye, & Tice, 1994). In a previous randomized controlled trial with families of adolescents with type 1 diabetes (Wysocki et al., 2000), 10 sessions of BFST over a 3-month interval yielded lasting improvements in parent–adolescent relationships and family communication skills as measured by parent and adolescent report (Wysocki et al., 2000) or by direct observation of structured family interactions (Wysocki et al., 1999). These benefits persisted for 12 months (Wysocki et al., 2001), and BFST was rated as more acceptable, applicable, and effective than was an educational support group (ES) (Wysocki et al., 1997). But, BFST did not impact glycemic control or treatment adherence (Wysocki et al., 2000, 2001). This second randomized trial, as reported here, sought to determine whether a revised intervention with diabetes-specific behavioral components, BFST for Diabetes (BFST-D), would have more robust effects on treatment adherence and diabetic control. Concurrently, various researchers reported the effectiveness of other intervention strategies with this clinical population (see Hampson et al., 2001 for a review). Drawing on this recent work, the BFST intervention was adapted by incorporating diabetes-specific elements of many of these approaches into BFST-D.
The following hypotheses were evaluated for this article:
- Hypothesis 1: Relative to SC and ES, families in BFST-D will exhibit significantly more improvement in parent–adolescent relationships (PARQ scores) and diabetes-related conflict (DRC scores) after 6 months of treatment.
- Hypothesis 2: Relative to SC and ES, adolescents in BFST-D will show significantly more improvement in medical adherence (DSMP scores) after 6 months of treatment.
- Hypothesis 3: Relative to SC and ES, adolescents in BFST-D will achieve significantly greater reduction in HbA1c, indicating better metabolic control, after 6 months of treatment.
- Hypothesis 4: Relative to SC and ES, adolescents in BFST-D will have significantly fewer hospitalizations and emergency room visits during the 6 months of study intervention.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionAcknowledgmentsReferences |
|---|
Participants and Settings
Recruitment occurred at two pediatric centers in the Southeastern and Midwestern United States. Recruitment objectives were to enroll a clinically appropriate sample of adolescents and their families who were experiencing significant problems with diabetes management. Parents and adolescents signed institutionally approved informed consent and assent forms, respectively, before any research procedures occurred.
Inclusion criteria were: adolescent age between 11 and 16 years inclusive; type 1 diabetes or insulin-treated type 2 diabetes for at least 2 years; HbA1c 8.0% (which has been defined as the threshold for clinical action by the American Diabetes Association, 2005); agreement to participate from all adult caregivers living with the adolescent; willingness to accept randomization; intent to continue diabetes care at the enrolling center for 18 months; and intent for the adolescent to remain living in the same home for 18 months. Exclusion criteria were: adolescent diagnosis of another systemic chronic disease except well-controlled asthma or Hashimoto’s thyroiditis; enrollment in self-contained special education; psychiatric admission of the adolescent within the prior 6 months; caregiver who was illiterate or not fluent in English; residence of adolescent in foster care, group home, or correctional facility; no telephone service; current diagnosis of psychosis, major depression, or substance abuse disorder in an adult caregiver; or open case with a child protection agency regarding child abuse or neglect.
Figure 1 provides a summary of participants’ progress through the various stages of the trial in the format recommended by the Consolidated Standards for Reporting Clinical Trials (CONSORT) (Begg et al., 1996). An introductory letter was mailed to 577 families, and telephone follow-up was achieved with 438 (76%) of them. Of these, 48 (11%) were determined to be ineligible by interview, while many others declined participation before their eligibility could be ascertained. A sample of 104 eligible families entered the study, 61 at the Southeastern site and 43 at the Midwestern site, for a recruitment rate of 27% of families who had not been found to be ineligible. Demographic characteristics of the sample are summarized in Table I and analyzed in the Results section.
|
|
Experimental Design
A three-group, randomized treatment design was used with four repeated measures at baseline, after treatment (6-months), and follow-up at 6 and 12 months after treatment. This article reports only the baseline and immediate posttreatment results as the sample continues in follow-up at this time. Following the baseline evaluation, as described below, families were randomized (stratified by baseline HbA1c) to standard care (SC), ES, or BFST-D for the next 6 months. The three experimental conditions are described below.
Standard Care (SC)
Diabetes care for all study participants reflected the prevailing clinical practices at each site during the study. Treating physicians selected an HbA1c target for each adolescent that was as close to the upper limit of normal (6.5%) as was considered safe and feasible. HbA1c was measured before each clinic visit and reviewed during the visit. Daily insulin replacement was achieved via multiple subcutaneous injections or insulin pump. Adolescents were asked to perform self-monitoring of blood glucose (SMBG) three or more times daily. Quarterly clinic visits were scheduled with a pediatric endocrinologist or other qualified clinician. A certified diabetes educator (CDE) provided basic and advanced diabetes education to families. Adolescents were offered a meal plan based on carbohydrate counting or an exchange system and encouraged to follow a personalized exercise plan. Adolescents and families were referred to qualified psychologists or psychiatrists not associated with the research team for services as needed.
Educational Support (ES)
In addition to the SC medical regimen, ES families attended 12 multifamily meetings within 6 months for diabetes education and social support. ES was designed to emulate a common mental health service for families of chronically ill teens and to serve as an alternative therapy comparison and a control for the differential professional attention received by the SC and BFST-D groups. Experienced diabetes nurses served as facilitators at each site and received extensive training before conducting ES sessions. Groups of three to five families completed a 12-session series together, attended by the parents and adolescents with diabetes. Session content followed the chapters of an American Diabetes Association curriculum for teens (Johnson, 2000). The same materials and session outlines were used at both sites, and the facilitators spoke weekly by telephone to ensure cross-site consistency. Family communication and conflict resolution skills were excluded from session content because these were specifically targeted by BFST-D. Sessions included a 45-min lecture by a health professional on 1 of the 12 topics, followed by 45 min of family interaction about that topic led by the facilitator.
Behavioral Family Systems Therapy for Diabetes (BFST-D)
In addition to the SC medical regimen, families in this group received 12 BFST-D sessions over 6 months. Sessions were conducted by one of three psychologists at the southeastern site or a licensed clinical social worker at the Midwestern site and were attended by the youth with diabetes and their caregivers who were participants. Therapists were trained and certified as proficient in BFST-D by two experienced, licensed psychologists (the first two authors) before enrollment of families. Methods described later in the article assured treatment integrity throughout the study.
BFST-D consisted of four components: problem-solving training, communication training, and cognitive restructuring and functional-structural family therapy. Problem-solving training provided families with a structured problem-solving approach with discrete steps consisting of: problem definition, generation of solutions, group decision making, planning, implementation and monitoring of the selected solution, and renegotiation or refinement of ineffective solutions. Communication skills training included instructions, feedback, modeling, and rehearsal targeting common parent–adolescent communication errors. Cognitive restructuring methods targeted family members’ irrational beliefs, attitudes, and attributions about one another’s behavior that could impede effective parent–adolescent communication. Functional and structural family therapy interventions targeted anomalous family systemic characteristics (e.g., weak parental coalitions and cross-generational coalitions) that could impede effective problem solving and communication. Families received the intervention components that were appropriate to their needs as determined by baseline assessments and ongoing observations in therapy. Sessions consisted of family problem solving and conflict resolution discussions. Therapists participated actively, frequently providing instructions, feedback, modeling, and rehearsal. Behavioral homework was assigned at each session and reviewed at the next session. Each session included delivery of some didactic information and emphasized teaching the family to acquire and apply the targeted skills at home. The revised BFST-D intervention utilized in this study included these diabetes-specific adaptations:
- All families targeted for treatment two or more diabetes-related problems that were identified as barriers to diabetes management or control. These behaviors were identified during the first two BFST-D sessions.
- Explicit training in behavioral contracting was provided (Carney, Schechter, & Davis, 1983; Epstein et al., 1981; Schafer, Glasgow & McCaul, 1982; Wysocki, Green & Huxtable, 1989). These studies have confirmed that behavioral contracting improves adolescents’ adherence to specific targeted aspects of the diabetes regimen. Behavioral contracts were evaluated and refined as needed in subsequent sessions.
- All families received advanced education in using SMBG data to modify insulin, diet, or exercise to improve diabetic control. The CDE and therapist conducted BFST-D sessions 6 and 7 jointly. The CDE provided education and training in using clinical algorithms for these purposes. The therapist helped families use their improved communication and problem-solving skills to facilitate this. Treatments with similar components have improved diabetic control (Anderson, Wolf, Burkhart, Cornell, & Bacon, 1989; Delamater et al., 1990).
- Parents simulated living with diabetes for 1 week between sessions 7 and 8. Satin, La Greca, Zigo, & Skyler (1989) showed that this intervention produced broad cognitive and affective benefits, and it is compatible with the BFST incorporation of cognitive restructuring methods. This included multiple daily injections of sterile saline on the child’s insulin schedule; daily blood glucose checks on the child’s schedule; monitoring and regulating carbohydrate intake; and managing one simulated hypoglycemic event (parents were notified by the therapist to indicate onset of "hypoglycemia").
- Therapists could extend the intervention to other social networks affecting the youths’ diabetes care by involving peers, siblings, teachers, or others and conducting sessions in other locations if necessary (Henggeler, Schoenwald, & Pickrel, 1995).
Participation Incentives
Participants were paid to promote adherence to the study tasks. Each family was paid $100 ($50 for parents and $50 for youth) for completing the scheduled evaluations. Each ES and BFST-D family received another $100, distributed in the same way, if they attended all 12 scheduled intervention sessions for their respective groups. These incentives resulted in >90% retention of the study sample in our previous work (Wysocki et al., 2001).
Measurement Methods and Schedule
Measures were collected at a baseline evaluation preceding randomization. Some measures were collected at baseline only for evaluation as moderators of treatment outcome, and those are not described in detail here since they were not included in the analyses for this article. These measures included the Brief Symptom Inventory (DeRogatis, 1977), Millon Adolescent Clinical Inventory (Millon, Millon, & Davis, 1997) and the Family Assessment Device (Kabacoff, Miller, Bishop, Epstein, & Keitner, 1990). Analyses pertinent to these measures will be the topic of a future report.
The General Information Form yielded demographic and medical information and the data that were needed to calculate the Hollingshead index of social status (1975, unpublished manuscript).
The following measures were collected at baseline and at the end of treatment (6 months):
The Parent–Adolescent Relationship Questionnaire (PARQ)
PARQ assesses the primary constructs of the behavioral family systems model (Robin, Koepke & Moye, 1990). It yields norm-referenced standard scores for three factor analytically derived scales: overt conflict/skill deficits, extreme beliefs, and family structure. There are separate forms for adolescents (314 true–false items) and parents (280 items). Internal consistency ranged from .73 to .89 for this sample. The proportions of study participants whose scores were at or above a clinical cut-off of T = 65 on "overt conflict/skill deficits" were 2.8% for adolescents, 2.0% for mothers, and 7.1% for fathers. For "extreme beliefs," the corresponding proportions of scores above the clinical cut-off were 9.2, 2.0, and 4.3%, respectively. Finally, for "family structure," clinically significant scores were obtained for 4.7, 3.8, and 1.7% of adolescents, mothers, and fathers, respectively.
The Diabetes Responsibility and Conflict Scale (DRC)
DRC assesses parent–child division of diabetes responsibilities and family conflict about 15 diabetes-related tasks (Rubin, Young-Hyman, & Peyrot, 1989). Internal consistency ranged from .86 to .92 for this sample.
The Diabetes Self-Management Profile (DSMP)
DSMP is a structured interview for the assessment of diabetes treatment adherence that was adapted from methods developed by Hanson, Henggeler, & Burghen (1987). The revised interview incorporates introductory comments designed to normalize the difficulty of regimen adherence. Questions were re-worded to enhance the ability to capture flexible diabetes self-management as it is currently practiced (e.g., insulin pump therapy). The revised interview (Harris et al., 2000) consisted of 24 adherence questions regarding 5 regimen domains (insulin, blood glucose testing, diet, exercise, and management of hypoglycemia). Internal consistency based on the present sample was .79 with 6-month test–retest reliability of .58. Scoring reliability across independent raters was .95. Total scores did not differ between parents and adolescents interviewed separately, and the correlation between parent and adolescent scores was r = .45 (p < .001). Baseline DSMP total scores correlated significantly with HbA1c levels with r = .29 (p < .01).
Glycosylated Hemoglobin (HbA1c)
HbA1c was measured quarterly at routine clinic visits with the DCA-2000 system (Bayer Diagnostics, Tarrytown, NY, USA). Assays completed at the two sites had been standardized against a reference laboratory, and the results confirmed its accuracy and consistency. The equivalence of DCA-2000 and reference laboratory measurements was also confirmed in a recent study (Diabetes Research in Children Network Study Group, 2005).
Assurance of Procedural Consistency and Integrity
Several methods ensured cross-site consistency of study procedures. A detailed operations manual described each study procedure. A previously prepared BFST manual was revised to reflect the BFST-D changes. The ES intervention was also guided by a manual (Johnson, 2000). Clinicians implementing the BFST-D and ES interventions spoke weekly by phone to plan intervention delivery. Additional efforts assured the integrity of BFST-D. A second therapist rated videotapes of each BFST-D session for treatment fidelity. The ratings were discussed in weekly phone conferences and deviations were resolved. The vast majority of ratings affirmed the integrity of the BFST-D intervention. Virtually all issues that were raised reflected legitimate differences in clinical judgment rather than violations of the written protocol.
Data Management and Statistical Analysis Plans
Research assistants checked all measures for appropriate completion before families left the evaluation session and corrected any administration errors. Data were then entered into a computer data file on a local area network at one of the centers. Data were checked for normality of distributions and for outliers or data entry errors before analyses. Statistical analyses began by comparing the groups at baseline on demographic and outcome variables. Since baseline HbA1c was correlated significantly with many of the outcome measures obtained at baseline, the general analysis plan for each outcome measure was to treat experimental groups (SC, ES, and BFST-D) and baseline HbA1c (above or below the median value of 9.0%) as between-subjects factors in repeated measures analyses of variance. The number of adolescents in each group with HbA1c <9.0% and 9.0%, respectively, was SC: 16 and 16; ES: 17 and 19; and BFST-D: 18 and 18. For analyses of treatment effects on HbA1c, the dependent variable was change in HbA1c from baseline to 6 months. To reduce the number of statistical comparisons, PARQ, DRC, and DSMP scores from adolescents and parents were combined to form family composite scores for these measures that were then submitted to statistical analyses. In all of these cases, family members’ scores were significantly correlated, with r-value ranging from .37 to .63.
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionAcknowledgmentsReferences |
|---|
Sampling and Randomization
Table I summarizes baseline demographic characteristics of the three groups. There were no statistically significant between-group differences at baseline on patient age, duration of diabetes, socioeconomic status, or of the distributions of gender, race/ethnicity, family composition, or parental marital status. The sample included the following proportion of families in each Hollingshead socioeconomic stratum: lower: 4.8%; lower middle: 21.2%; middle: 38.5%; upper middle: 24.0%; and upper: 11.5%. The groups also did not differ significantly at baseline with respect to HbA1c or scores on the PARQ, DRC, or DSMP. Members of racial and ethnic minorities comprised 37% of the sample. Among the 104 families who were randomized, 92 (29 SC, 35 ES, and 28 BFST-D) completed the baseline and 6-month evaluations reported in this article. The 12 dropouts did not differ significantly at baseline from those who completed the study in terms of any outcome measure analyzed for this paper. Compared to those who completed the study, dropouts had significantly lower SES (mean Hollingshead index = 41.5 versus 30.7) and were significantly less likely to be living with both biological parents (45 versus 25%). There were no statistically significant differences across sites or therapists on any of the outcome measures at baseline or 6-month follow-up.
Hypothesis 1 (Family Relationships and Conflict)
Table II summarizes that change in PARQ family composite T scores from baseline to 6 months did not differ significantly among treatment groups for the overt conflict/skill deficits, extreme beliefs or family structure subscales. Neither the main effect for group nor the group x time interaction effect was statistically significant. Hypothesis 1 was therefore not confirmed relative to effects on the PARQ.
|
Mean DRC family composite scores for each group at baseline and 6 months, respectively, were SC: 25.9 and 29.4; ES: 29.4 and 32.1; and BFST-D: 27.5 and 26.9. The increase in DRC scores in the SC and ES groups and slight decrease for the BFST-D group yielded a significant group x time interaction effect [F(2,89) = 4.31; p <.02]. Further, a significant group x time–baseline HbA1c interaction [F(2,87) = 3.27; p <.04] indicated that this treatment effect varied in magnitude as a function of baseline HbA1c level. Figure 1 portrays this interaction graphically by presenting change in family composite DRC scores from baseline to 6 months for youths with HbA1c above and below the median baseline level (9.0%) for each group. BFST-D yielded significantly greater reduction in DRC scores than either of the other two groups, and this effect was limited to those with baseline HbA1c above 9.0%. Hypothesis 1 was therefore confirmed with respect to differential treatment effects on DRC scores, although this effect occurred only for those in the poorest diabetic control at baseline, and the effect is best described as attenuation by BFST-D of worsening conflict that occurred among the comparison groups.
Hypothesis 2 (Treatment Adherence)
Mean treatment adherence (DSMP) scores obtained for each group at baseline and 6 months, respectively, were SC: 53.0 and 51.4; ES: 55.2 and 54.6; and BFST-D: 54.7 and 57.4. The main effect for groups and the group x time interaction effect failed to achieve statistical significance. However, there was a significant group x time x baseline HbA1c interaction effect [F(2,89) = 3.63; p <.03]. Figure 2 illustrates change in DSMP total scores from baseline to 6 months as a function of treatment group and baseline HbA1c. Post-hoc analyses showed that BFST-D yielded significantly greater improvement (or less decline) in treatment adherence relative to both of the other treatment groups and within both baseline HbA1c ranges. Hypothesis 2 was therefore confirmed with respect to superiority of BFST-D over the other treatment groups in terms of effects on treatment adherence.
|
Hypothesis 3 (Glycemic Control)
Mean HbA1c levels at baseline and 6 months, respectively, were SC: 9.5 and 9.2%; ES: 9.7 and 8.9%, and BFST-D: 9.6 and 8.8%. Neither the main effect for groups nor the group x time interaction effect were significant. But, as illustrated in Fig. 4, the group x time x baseline HbA1c interaction effect were statistically significant [F(2,89) = 3.17; p <.05]. Although HbA1c tended to decline somewhat for all three treatment groups, post-hoc comparisons showed that, among those with baseline HbA1c above 9.0%, improvement in HbA1c was significantly greater for both the BFST-D (–1.3%) and the ES groups (–1.1%) than for the SC group (–0.4%). Figure 3 portrays the interactive effects of the treatments on change in HbA1c as a function of baseline HbA1c. The difference between the ES and BFST-D groups did not achieve statistical significance. Hypothesis 3 was therefore partially confirmed in that BFST-D was superior to the SC group, but not the ES group, in terms of improving metabolic control during the study. The benefits of both ES and BFST-D were most evident for those with poor metabolic control at baseline (Fig. 4).
|
|
Hypothesis 4 (Health Care Utilization)
Participants had 10 hospitalizations and 12 emergency room visits during the 6-month study. Hospital admissions for each group were SC: 4; ES: 3; and BFST-D: 3. Emergency room visits for each group was SC: 4; ES: 5; and BFST-D: 3. Inferential statistical analysis was inappropriate due to the very low frequencies of these events and because 13 of the 22 episodes were not related directly to diabetes.
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionAcknowledgmentsReferences |
|---|
This article reports the immediate posttreatment results of a randomized controlled trial comparing SC for diabetes mellitus alone to augmentation of SC by 6 months of family participation in either a 12-session multifamily ES or to 12 sessions of BFST-D. The BFST-D treatment had been revised in response to a previous trial in an effort to enhance the effects of the intervention on treatment adherence, glycemic control, and health care utilization (Wysocki et al., 1997
, 1999, 2000, 2001). Effectiveness of this revised BFST-D intervention relative to the SC and ES conditions was evaluated by comparing change in four sets of outcome measures over 6 months of treatment. Overall, the results yielded support for the hypothesized effectiveness of BFST-D and suggested that the refinements made to the previously tested intervention did enhance the intervention’s impact on certain diabetes outcomes. This adds to a growing literature describing other empirically validated interventions for this population (Anderson et al., 1989; Delamater et al., 1990; Ellis et al., 2005; Ellis, Naar-King, Frey, Rowland & Greger, 2003; Grey, Boland, Davidson, Li, & Tamborlane, 2000; Hampson et al., 2001; Laffel et al., 2003) by showing that BFST-D yielded significant benefits to adolescents and their families in terms of several diabetes-specific outcomes. Hypothesis 1 predicted that BFST-D would yield significantly greater improvement in parent–adolescent relationships (PARQ scores) and diabetes-related family conflict (DRC scores) compared to SC or ES. Hypothesis 1 was partially confirmed since significant group x time and group x time x baseline HbA1c interactions were obtained. Diabetes-related family conflict was decreased slightly during treatment for the BFST-D group, whereas it increased for the SC and ES groups during the same period. This difference, as illustrated in Fig. 2, was attributable primarily to fairly substantial treatment effects for youths with the poorest baseline HbA1c levels and is best described as prevention of worsening of diabetes-related family conflict that occurred in the SC and ES groups. The absence of corresponding effects on more general parent–adolescent relationships (i.e., PARQ scores) may reflect the emphasis within BFST-D on targeting diabetes-specific problems rather than focusing on more general aspects of family relationships. Also, in distinction to the present trial, the enrollment criteria for the previous BFST trial required that families exceeded specific cut-off scores indicating moderate or greater family conflict. Consequently, PARQ scores for the present sample were comparatively lower, possibly diminishing the capacity to detect treatment effects on this measure. Alternatively, study participation may have sensitized some participants to problems in family communication or conflict resolution that were not readily apparent to them at baseline. Each of these interpretations is quite speculative at this time.
Hypothesis 2 consisted of the prediction that BFST-D would yield more improvement in diabetes self-management behaviors (DSMP scores) compared with either SC or ES. The results clearly confirmed Hypothesis 2 by revealing a statistically significant main effect for treatment group favoring BFST over SC and ES and by a significant group x time interaction effect, as shown in Fig. 3. DSMP scores were significantly more favorable for the BFST-D group than the SC or ES groups within both baseline HbA1c ranges.
Hypothesis 3 addressed the comparative effects of SC, ES and BFST-D on change in HbA1c levels during treatment. As illustrated in Fig. 4, both ES and BFST-D yielded substantial reductions in HbA1c among those with baseline levels above 9.0%. The magnitude of these changes is considered to be quite significant clinically because these reductions amount to an approximate 0.7 standard deviation difference relative to baseline for the ES group and a 0.8 standard deviation difference for the BFST-D group.
Hypothesis 4 was not submitted to statistical analysis because the frequencies of hospitalizations and emergency room visits were so low. The recorded 10 hospitalizations and 12 emergency room visits were distributed evenly across groups, and most of these events were not diabetes related. Further follow-up of the sample for an additional year may clarify whether these measures of health care utilization differ among the treatment groups over a longer interval.
Overall, the study results suggest substantial promise for the revised BFST-D intervention for reducing family conflict about diabetes and for improving treatment adherence and metabolic control. The observation that these effects were more evident for families of those youths entering the study with poor metabolic control is encouraging, because these are the youths who often endure prolonged maladjustment to diabetes and a downward spiral of poor metabolic control, psychiatric disorders, acceleration of long-term complications, and excess health care costs (Bryden et al., 2001; Wysocki, Hough, Ward & Green, 1992).
Mechanisms that may account for the stronger BFST-D treatment effects on HbA1c and diabetes-related family conflict (DRC scores) among families of adolescents with poor metabolic control at baseline remain to be identified. Possibly, specification of treatment targets with such adolescents and their families may be more obvious than for those whose diabetes-management difficulties are somewhat less salient. When families exhibit glaring problems with diabetes management, modest behavior change may yield more pronounced benefits that might be less evident for families with slightly more competence. Stated another way, adolescents in very poor metabolic control before treatment may simply have had more room for improvement in both self-management behaviors and glycemic control. Whether these or other explanations prove accurate, it is encouraging that adolescents with poor metabolic control enjoyed clinically meaningful treatment gains through BFST-D.
Certain limitations of the study should be noted. The feasibility of delivery of BFST-D in most clinical settings, especially given the constraints imposed by the health care economy, is questionable. The study was conducted under optimal circumstances that are unlikely to exist in typical clinical settings. Participants were paid for completing study evaluations, intervention sessions were provided without charge, and substantial flexibility in appointment times was offered to make engagement in therapy as convenient as possible. These features are unlikely to be feasible in most pediatric settings. Future research on BFST-D should focus on optimizing its cost effectiveness through such mechanisms as reducing the number of sessions, evaluating delivery to multiple families concurrently rather than to individual families, or supplementation of BFST-D sessions via internet or interactive telecommunication methods.
Other important limitations relate to sampling and retention. The sample size was relatively small for a trial of this type, and 88% of the sample completed both evaluations. As a consequence, the cell sizes for certain of the interaction effects reported in this article were sometimes small, increasing the likelihood that spurious effects would prove to be statistically significant. However, the benefits of BFST-D to those with poor metabolic control were confirmed across three different outcome measures including questionnaires (DRC), structured interviews (DSMP), and biochemical assays (HbA1c). The consistency of these findings across different outcome measures and methods bolsters confidence that these were not spurious results.
Also, the selective benefits from BFST-D accruing to those with poor baseline metabolic control were identified through secondary data analyses and were not stated as a priori scientific hypothesis. Confidence in these findings and the conclusions drawn from them would be stronger had these observed relationships been hypothesized in advance.
Finally, the present article evaluated a complex intervention with multiple components. Consequently, "dismantling studies" to identify the critical elements of BFST-D may prove difficult.
The present article addresses only the immediate treatment effects of BFST-D on the various outcomes. Although the previous trial (Wysocki et al., 1997, 1999, 2000, 2001) demonstrated lasting treatment effects on measures of parent–-adolescent relationships, diabetes-related family conflict, and treatment satisfaction, the durability of the treatment effects reported here remains to be confirmed. Follow-up of the present sample for an additional 12 months after treatment will be the topic of subsequent reports.
|
Acknowledgments |
|---|
|
TopAbstractMethodsResultsDiscussionAcknowledgmentsReferences |
|---|
This study was supported by NIH grant 1-RO1-DK43802 to the first author and NIH grants P60-DK20579 and RR00036 which support the Diabetes Research and Training Center and General Clinical Research Center at the Washington University School of Medicine.
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionAcknowledgmentsReferences |
|---|
American Diabetes Association. (2005). Clinical Practice Recommendations, 2005. Standards of Medical Care in Diabetes. Diabetes Care, 28(Suppl. 1), S4–S36.
Anderson, B. J., Miller, J. P., Auslander, W. F., & Santiago, J. V. (1981). Family characteristics of diabetic adolescents: Relationships to metabolic control. Diabetes Care, 4, 586–594.[Abstract]
Anderson, B. J., Wolf, F. M., Burkhart, M. T., Cornell, R. G., & Bacon, G. E. (1989). Effects of a peer group intervention on metabolic control of adolescents with IDDM: Randomized outpatient study. Diabetes Care, 12, 184–188.[Abstract]
Barkley, R., Guevremont, D., Anastopoulos, A., & Fletcher, K. (1992). A comparison of three family therapy programs for treating family conflict in adolescents with attention-deficit hyperactivity disorder. Journal of Consulting and Clinical Psychology, 60, 450–462.[CrossRef][ISI][Medline]
Begg, C., Cho, M., Eastwood, S., Horton, R., Moher, D., Olkin, I., Pitkin, R., Rennie, D., Schulz, K. F., Simel, D., & Stroup, D. F. (1996). Improving the quality of reporting of randomized controlled trials: the CONSORT statement. Journal of the American Medical Association, 276, 637–639.[CrossRef][ISI][Medline]
Bobrow, E. S., AvRuskin, T. W., & Siller, J. (1985). Mother–daughter interaction and adherence to diabetes regimens. Diabetes Care, 8, 146–151.[Abstract]
Bryden, K. S., Peveler, R. C., Stein, A., Neil, A., Mayou, R. A., & Dunger, D. B. (2001). Clinical and psychological course of diabetes from adolescence to young adulthood: a longitudinal cohort study. Diabetes Care, 24, 1536–1540.
Carney, R. M., Schechter, K., & Davis, T. (1983). Improving adherence to blood glucose monitoring in insulin-dependent diabetic children. Behavior Therapy, 14, 247–254.
Delamater, A. M., Davis, S., Bubb, J., Smith, J., Schmidt, L., White, N. H., et al. (1990). Randomized prospective study of self management training with newly diagnosed diabetic children. Diabetes Care, 13, 492–498.[Abstract]
DeRogatis, R. (1977). Symptom Checklist 90R: Administration and Scoring Manual. Baltimore, MD: Clinical Psychometrics Research.
Diabetes Research in Children Network Study Group. (2005). Performance of the DCA2000 for measurement of HbA1c levels in children with T1DM in a DirecNet outpatient clinical trial. Pediatric Diabetes, 6, 13–16.[CrossRef][ISI][Medline]
Ellis, D. A., Frey, M., Naar-King, S., Templin, T., Cunningham, P., & Cakan, N. (2005). Use of multisystemic therapy to improve regimen adherence among adolescents with type 1 diabetes in chronic poor metabolic control: a randomized controlled trial. Diabetes Care, 28, 1604–1610.
Ellis, D. A., Naar-King, S., Frey, M., Rowland, M., & Greger, N. (2003). Case study: feasibility of multisystemic therapy as a treatment for urban adolescents with poorly controlled type 1 diabetes. Journal of Pediatric Psychology, 28, 287–293.
Epstein, L. H., Beck, S., Figueroa, J., Farkas, G., Kazdin, A. E., Daneman, D., et al. (1981). The effects of targeting improvement in urine glucose on metabolic control in children with insulin-dependent diabetes mellitus. Journal of Applied Behavior Analysis, 14, 365–375.[CrossRef][ISI][Medline]
Foster, S. L., Prinz, R. J., & O’Leary, K. D. (1983). Impact of problem-solving communication training and generalization procedures on family conflict. Child and Family Behavior Therapy, 5, 1–23.
Grey, M., Boland, E., Davidson, M., Li, J., & Tamborlane, W. V. (2000). Coping skills training for youth on intensive therapy has long-lasting effects on metabolic control and quality of life. Journal of Pediatrics, 137, 107–113.[CrossRef][ISI][Medline]
Gustafsson, P., Cederblad, M., Ludvigsson, J., & Lundin, B. (1987). Family interaction and metabolic balance in juvenile diabetes mellitus: a prospective study. Diabetes Research and Clinical Practice, 4, 7–14.[CrossRef][ISI][Medline]
Hampson, S. E., Skinner, T. C., Hart, J., Storey, L., Gage, H., Foxcroft, D., et al. (2001). Effects of educational and psychological interventions for adolescents with diabetes mellitus: a systematic review. Health Technology Assessment, 5, 1–77.
Hanson, C., Henggeler, S., & Burghen, G. (1987). Social competence and parental support as mediators of the link between stress and metabolic control in adolescents with insulin-dependent diabetes mellitus. Journal of Consulting and Clinical Psychology, 55, 529–533.[CrossRef][ISI][Medline]
Harris, M. A., Wysocki, T., Sadler, M., Wilkinson, K., Harvey, L. M., Buckloh, L. M., et al. (2000). Validation of a structured interview for the assessment of diabetes self management. Diabetes Care, 23, 1301–1304.
Hauser, S. T., Jacobson, A. M., Lavori, P., Wolfdorf, J. I., Herskowitz, R. D., Milley, J. E., et al. (1990). Adherence among children and adolescents with insulin-dependent diabetes mellitus over a four year follow-up. II: immediate and long-term linkages with the family milieu. Journal of Pediatric Psychology, 15, 527–542.
Henggeler, S. W., Schoenwald, S. K., & Pickrel, S. G. (1995). Multisystemic therapy: Bridging the gap between University- and community-based treatment. Journal of Consulting and Clinical Psychology, 63, 709–717.[CrossRef][ISI][Medline]
Hollingshead, A. B. (1975). A four factor index of social status. Unpublished Manuscript, Yale University.
Johnson, D. (2000). Teenagers with type 1 diabetes: a curriculum for adolescents and families. Alexandria, VA: American Diabetes Association.
Kabacoff, R. I., Miller, I. W., Bishop, D. S., Epstein, N. B., & Keitner, G. I. (1990). A psychometric study of the McMaster Family Assessment Device in psychiatric, medical and non-clinical samples. Journal of Family Psychology, 3, 431–439.[CrossRef]
Koski, M., Ahlas, A., & Kumento, A. (1976). A psychosomatic follow-up study of juvenile diabetics. Acta Paedopsychiatrica, 42, 12–25.[ISI][Medline]
Laffel, L., Vangsness, L., Connell, A., Goebel-Fabbri, A., Butler, D., & Anderson, B. J. (2003). Impact of ambulatory, family-focused teamwork intervention on glycemic control in youth with type 1 diabetes. Journal of Pediatrics, 142, 409–416.[CrossRef][ISI][Medline]
Miller-Johnson, S., Emery, R. E., Marvin, R. S., Clarke, W., Lovinger, R., & Martin, M. (1994). Parent–child relationships and the management of insulin-dependent diabetes mellitus. Journal of Consulting and Clinical Psychology, 62, 603–610.[CrossRef][ISI][Medline]
Millon, T., Millon, C., & Davis, R. (1997). Millon Adolescent Clinical Inventory Manual. Minneapolis, MN: NCS Assessments.
Robin, A. L., & Foster, S. L. (1989). Negotiating parent–adolescent conflict: a behavioral family systems approach. New York: Guilford Press.
Robin, A. L., Koepke, T., & Moye, A. W. (1990). Multidimensional assessment of parent–adolescent relations. Psychological Assessment, 10, 451–459.
Robin, A. L., Siegel, P. T., Koepke, T., Moye, A. W., & Tice, S. (1994). Family therapy versus individual therapy for adolescent females with anorexia nervosa. Developmental and Behavioral Pediatrics, 15, 111–116.
Rubin, R. R., Young-Hyman, D., & Peyrot, M. (1989). Parent–child responsibility and conflict in diabetes care [Abstract]. Diabetes, 38(Suppl. 2), 28A.
Satin, W., La Greca, A. M., Zigo, S., & Skyler, J. S. (1989). Diabetes in adolescence: Effects of multifamily group intervention and parent simulation of diabetes. Journal of Pediatric Psychology, 14, 259–276.
Schafer, L. C., Glasgow, R. E., & McCaul, K. D. (1982). Increasing the adherence of diabetic adolescents. Journal of Behavioral Medicine, 5, 353–362.[CrossRef][ISI][Medline]
Wysocki, T. (1993). Associations among parent–adolescent relationships, metabolic control and adjustment to diabetes in adolescence. Journal of Pediatric Psychology, 18, 443–454.
Wysocki, T., Greco, P., Harris, M. A., Bubb, J., & White, N. H. (2001). Behavior therapy for families of adolescents with diabetes: Maintenance of treatment effects. Diabetes Care, 24, 441–446.
Wysocki, T., Green, L. B., & Huxtable, K. (1989). Blood glucose monitoring by diabetic adolescents: Compliance and metabolic control. Health Psychology, 8, 267–284.[CrossRef][ISI][Medline]
Wysocki, T., Harris, M. A., Greco, P., Bubb, J., Elder, C. L., et al. (2000). Randomized controlled trial of behavior therapy for families of adolescents with insulin-dependent diabetes mellitus. Journal of Pediatric Psychology, 25, 23–33.
Wysocki, T., Harris, M., Greco, P., Harvey, L., McDonell, K., Elder, C., et al. (1997). Social validity of support group and behavior therapy interventions for families of adolescents with insulin-dependent diabetes mellitus. Journal of Pediatric Psychology, 22, 635–649.
Wysocki, T., Hough, B. S., Ward, K. M., & Green, L. B. (1992). Diabetes mellitus in the transition to adulthood: Adjustment, self-care and health status. Journal of Developmental and Behavioral Pediatrics, 13, 194–201.[ISI][Medline]
Wysocki, T., Miller, K., Greco, P., Harris, M. A., Harvey, L., Taylor, A., et al. (1999). Behavior therapy for families of adolescents with diabetes: Effects on directly observed family interactions. Behavior Therapy, 30, 496–515.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. C. Duke, G. R. Geffken, A. B. Lewin, L. B. Williams, E. A. Storch, and J. H. Silverstein Glycemic Control in Youth with Type 1 Diabetes: Family Predictors and Mediators J. Pediatr. Psychol., August 1, 2008; 33(7): 719 - 727. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kahana, D. Drotar, and T. Frazier Meta-Analysis of Psychological Interventions to Promote Adherence to Treatment in Pediatric Chronic Health Conditions J. Pediatr. Psychol., July 1, 2008; 33(6): 590 - 611. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Wysocki Editorial: Journal of Pediatric Psychology Statement of Purpose Section on Randomized Trials J. Pediatr. Psychol., January 1, 2008; 33(1): 12 - 15. [Full Text] [PDF]
|
|
|
|
|
|
L. P. Barakat Editorial: Journal of Pediatric Psychology Statement of Purpose Section on Family Influences and Adaptation J. Pediatr. Psychol., January 1, 2008; 33(1): 26 - 30. [Full Text] [PDF]
|
|
|
|
|
|
J. M. Butler, M. Skinner, D. Gelfand, C. A. Berg, and D. J. Wiebe Maternal Parenting Style and Adjustment in Adolescents with Type I Diabetes J. Pediatr. Psychol., November 1, 2007; 32(10): 1227 - 1237. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. K. Hood and T. R. Nansel Commonalities in Effective Behavioral Interventions for Children and Adolescents With Type 1 Diabetes: A Review of Reviews Diabetes Spectr, October 1, 2007; 20(4): 251 - 254. [Full Text] [PDF]
|
|
|
|
 |
|
 M. Peyrot and R. R. Rubin Behavioral and Psychosocial Interventions in Diabetes: A conceptual review Diabetes Care, October 1, 2007; 30(10): 2433 - 2440. [Full Text] [PDF]
|
|
|
|
|
|
K. K. Hood, D. A. Butler, B. J. Anderson, and L. M.B. Laffel Updated and Revised Diabetes Family Conflict Scale Diabetes Care, July 1, 2007; 30(7): 1764 - 1769. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Wysocki, M. A. Harris, L. M. Buckloh, D. Mertlich, A. S. Lochrie, N. Mauras, and N. H. White Randomized Trial of Behavioral Family Systems Therapy for Diabetes: Maintenance of effects on diabetes outcomes in adolescents Diabetes Care, March 1, 2007; 30(3): 555 - 560. [Abstract] [Full Text] [PDF]
|
|
| ||||||||